mapFields with moving meshes

Bazinga · February 22, 2019, 07:07

I am facing a problem where I have a moving body with sixDoFRigidBodyMotion and I want to check the sensitivity of the results regarding coarser/finer meshes. When I initialize on a different mesh, the location of the body and the mesh deformation information is not mapped to the new field.

Does anyone have a solution to this problem?

simrego · February 22, 2019, 09:20

Hi!

Is it even possible? With dynamic mesh you have a "new mesh" at every step. Actually new points if there are no topological changes.
I think you can only map fields but not the actual point locations of the mesh. Your mesh is distorted. I think in most of the cases it is nearly impossible. For example you have a wing which is moved upwards. Then you try to map these mesh points into a new mesh, but there will be many points without overlapping and it is a nightmare. Also you have no idea which point was which. Even if you map field from one mesh to another and if the domain is not the same the mapping algorithm can do interesting things especially where you should map data from outside of the source domain.

What if you just run the solver with local time stepping (if the solver has this functionality) from the beginning to reach the steady state solution and then run the pure transient part.

February 22, 2019, 07:07	mapFields with moving meshes	#1
Bazinga Senior Member Join Date: Jun 2012 Location: Germany, Bochum Posts: 230 Rep Power: 16	I am facing a problem where I have a moving body with sixDoFRigidBodyMotion and I want to check the sensitivity of the results regarding coarser/finer meshes. When I initialize on a different mesh, the location of the body and the mesh deformation information is not mapped to the new field. Does anyone have a solution to this problem?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Issues with mapFields	BlackBoatNavArch	OpenFOAM Pre-Processing	38	May 28, 2021 17:29
Moving meshes: Arbitrary Lagrangian-Eulerian (ALE) formulation	Joachim	Main CFD Forum	0	August 29, 2014 12:03
Confirmation Needed on fvMeshSubset and Moving Meshes	ngj	OpenFOAM Programming & Development	1	September 17, 2010 09:06
Problems with mapFields and moving mesh.	ghindi	OpenFOAM Pre-Processing	2	April 25, 2010 06:56
Moving mesh in Fluent	fivos	FLUENT	0	April 2, 2010 10:45

February 22, 2019, 09:20		#2
simrego Senior Member anonymous Join Date: Jan 2016 Posts: 416 Rep Power: 14	Hi! Is it even possible? With dynamic mesh you have a "new mesh" at every step. Actually new points if there are no topological changes. I think you can only map fields but not the actual point locations of the mesh. Your mesh is distorted. I think in most of the cases it is nearly impossible. For example you have a wing which is moved upwards. Then you try to map these mesh points into a new mesh, but there will be many points without overlapping and it is a nightmare. Also you have no idea which point was which. Even if you map field from one mesh to another and if the domain is not the same the mapping algorithm can do interesting things especially where you should map data from outside of the source domain. What if you just run the solver with local time stepping (if the solver has this functionality) from the beginning to reach the steady state solution and then run the pure transient part.